A precisely positioned MED12 activation helix stimulates CDK8 kinase activity
Felix Klatta, Alexander Leitnerb, Iana V. Kima, Hung Ho-Xuanc, Elisabeth V. Schneiderd,e, Franziska Langhammera, Robin Weinmanna, Melanie R. Müllera, Robert Huberd,f,g,1, Gunter Meisterc, and Claus-D. Kuhna,1
aGene Regulation by Non-Coding RNA, Elite Network of Bavaria and University of Bayreuth, 95447 Bayreuth, Germany; bDepartment of Biology, Institute of Molecular Systems Biology, Eidgenössiche Technische Hochschule Zürich, 8093 Zürich, Switzerland; cBiochemistry Center Regensburg, Laboratory for RNA Biology, University of Regensburg, 93053 Regensburg, Germany; dMax Planck Institute of Biochemistry, 82152 Martinsried, Germany; eProteros Biostructures GmbH, 82152 Martinsried, Germany; fCenter of Medical Biotechnology, University of Duisburg-Essen, 45117 Essen, Germany; and gDepartment of Chemistry, Technical University of Munich, 85747 Garching, Germany
Contributed by Robert Huber, December 21, 2019 (sent for review October 10, 2019; reviewed by Patrick Cramer and Matthias Geyer) The Mediator kinase module regulates eukaryotic transcription by with altered subunit composition accompanied by a potentially phosphorylating transcription-related targets and by modulating altered activity and target specificity (14). CDK8 and its paralog the association of Mediator and RNA polymerase II. The activity of its CDK19 were found to phosphorylate a wide variety of substrates catalytic core, cyclin-dependent kinase 8 (CDK8), is controlled by involved in transcription, DNA repair, and metabolic processes Cyclin C and regulatory subunit MED12, with its deregulation (11). This wealth of substrates likely lies at the heart of the context- contributing to numerous malignancies. Here, we combine in vitro dependent role of the Mediator kinase module in metazoan tran- biochemistry, cross-linking coupled to mass spectrometry, and scription. Moreover, the central role of CDK8 in transcription in vivo studies to describe the binding location of the N-terminal regulation explains why it is a potent oncogene that is involved in segment of MED12 on the CDK8/Cyclin C complex and to gain the progression of malignant cancers, like colorectal cancer and mechanistic insights into the activation of CDK8 by MED12. Our data acute myeloid leukemia (15–17). Due to its oncogenic potential, demonstrate that the N-terminal portion of MED12 wraps around the effective and selective inhibition of CDK8 is of great interest to CDK8, whereby it positions an “activation helix” close to the T-loop drug development and therefore spurred significant efforts to de- of CDK8 for its activation. Intriguingly, mutations in the activation velop CDK8-specific inhibitors. Despite the availability of structural
helix that are frequently found in cancers do not diminish the affinity information on CDK8 bound to Cyclin C (18), the development of BIOCHEMISTRY of MED12 for CDK8, yet likely alter the exact positioning of the CDK8-specific inhibitors is hampered by the to date unclear activation helix. Furthermore, we find the transcriptome-wide mechanism of CDK8 activation. In contrast to the common acti- gene-expression changes in human cells that result from a mutation vation mechanism of cyclin-dependent kinases, CDK8 does not in the MED12 activation helix to correlate with deregulated genes in require phosphorylation of its T-loop for full activity. Instead, breast and colon cancer. Finally, functional assays in the presence of CDK8 T-loop phosphorylation seems to be functionally replaced by kinase inhibitors reveal that binding of MED12 remodels the active MED12 binding to CDK8, which results in its activation (5, 19). site of CDK8 and thereby precludes the inhibition of ternary CDK8 complexes by type II kinase inhibitors. Taken together, our results Significance not only allow us to propose a revised model of how CDK8 activity is regulated by MED12, but also offer a path forward in developing small molecules that target CDK8 in its MED12-bound form. CDK8, the catalytic core of the Mediator kinase module, is activated by MED12. We describe the architecture of the ternary complex of CDK8 | MED12 | Mediator kinase module | Mediator | kinase inhibitors CDK8, Cyclin C, and MED12, and we decipher the mechanism by which MED12 activates CDK8. We find the N-terminal segment of MED12towraparoundtheCDK8moleculetoplacean“activation ediator is essential for all RNA polymerase II (Pol II)- helix” in proximity of its T-loop. Exact placement of the helix, not Mmediated transcription in eukaryotes because it allows for MED12 binding alone, is crucial for CDK8 activation. Moreover, the communication between enhancer-bound transcription factors – MED12 binding precludes binding of type II kinase inhibitors to the and Pol II (1 4). In humans Mediator comprises 30 protein subunits active site of CDK8. As many cancer patients carry mutations in the that are subdivided into four modules: The head, the middle, the tail, MED12 activation helix, our study invigorates the development of and the kinase module. Whereas the first three modules form a CDK8-specific compounds as well as offers functional insights into complex that is often referred to as the Mediator complex, the kinase the Mediator kinase module. module only reversibly associates with the three-module Mediator – complex (5 7). Structural information on the isolated Mediator Author contributions: F.K. and C.-D.K. designed research; F.K., A.L., I.V.K., H.H.-X., E.V.S., complex and on the head and the middle module of Mediator, F.L., R.W., and M.R.M. performed research; G.M. and C.-D.K. supervised the study; F.K., collectively termed core Mediator, bound to Pol II provided first A.L., I.V.K., E.V.S., and R.H. analyzed data; and F.K., R.H., G.M., and C.-D.K. wrote insights into how Mediator is able to activate Pol II (8, 9). In contrast the paper. to the Mediator complex, the structure of the kinase module is un- Reviewers: P.C., Max Planck Institute for Biophysical Chemistry; and M.G., University known and its role during the transcription cycle is unclear. Despite of Bonn. the fact that, in yeast, the three-module Mediator complex is unable The authors declare no competing interest. to simultaneously bind Pol II and the kinase module (10), deletion or Published under the PNAS license. inhibitor of CDK8 results only in minor gene-expression changes (11, Data deposition: The mass spectrometry data were deposited to the ProteomeXchange 12). However, using human proteins in a mostly recombinant in vitro Consortium via the PRIDE partner repository (identifier PXD015394). The structure of Compound A in complex with CDK8 (1–403)/Cyclin C were deposited to the Protein Data system, the kinase module was found to repress transcription by Bank (PDB ID code 6T41). All sequencing data been deposited in the Gene Expression inhibiting Mediator coactivator function (13). Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo (accession no. GSE135458). The Mediator kinase module consists of four subunits: Cyclin- 1To whom correspondence may be addressed. Email: [email protected] or claus. dependent kinase 8 (CDK8), its corresponding cyclin, Cyclin C, and [email protected]. two large regulatory subunits, MED12 and MED13. Moreover, in This article contains supporting information online at https://www.pnas.org/lookup/suppl/ metazoans all kinase module subunits, except for Cyclin C, possess doi:10.1073/pnas.1917635117/-/DCSupplemental. paralogues that can be utilized for the assembly of kinase modules
www.pnas.org/cgi/doi/10.1073/pnas.1917635117 PNAS Latest Articles | 1of12 Downloaded by guest on September 27, 2021 The current model for CDK8 activation proposes that the N-terminal necessary for CDK8 activation. In addition, we find that disease- segmentofMED12bindstoasurfacegrooveonCyclinC,andthereby causing MED12 mutations likely lead to mis-positioning of the enhances CDK8 activity (19). However, how MED12 binding to activation helix, and thereby abrogate CDK8 activity. In further this distant surface groove could be coupled to repositioning of support, we show that the transcriptome-wide gene expression the CDK8 T-loop remains unclear and the current model is changes in HCT116 cells that carry a MED12 E33Q mutation therefore controversial. phenocopy multiple cancer-related transcriptome alterations. Fi- To investigate how MED12 activates CDK8, we established a nally, we demonstrate that, in contrast to type I inhibitors, type II recombinant expression and purification scheme that resulted in kinase inhibitors lose much of their inhibitory potential when used monodisperse, highly pure ternary CDK8/Cyclin C/MED12 against ternary MED12-bound CDK8/Cyclin C in vitro. Taken complexes. Using these complexes, we carried out chemical cross- together, our results establish that an activation helix in MED12 linking coupled to mass spectrometry. Unexpectedly, the resulting functionally replaces T-loop phosphorylation of CDK8. Further- lysine- and aspartate/glutamate-mediated cross-links suggest a more, MED12 binding likely remodels the CDK8 active site, thus substantially revised mechanism for how the N-terminal portion of preventing type II kinase inhibitors from efficiently targeting MED12 binds and activates CDK8: MED12 makes extensive CDK8 bound by MED12 in vivo. contacts to CDK8 with no crucial contacts to Cyclin C. In- triguingly, MED12 thereby places a critical “activation helix” in Results direct vicinity of the T-loop of CDK8, leading to its conformational MED12 Binding Stabilizes and Stimulates the Activity of the Binary stabilization by contacting its arginine triad. Using kinase assays CDK8/Cyclin C Complex. To study the affinity of MED12 for the and mutational profiling, we demonstrate that the N-terminal tip binary CDK8/Cyclin C complex, we devised an expression and of the MED12 activation helix, in particular glutamate-33 (E33), is purification strategy for both components in insect cells (Fig. 1A
CDK8 CDK8 CDK8 MED12 (1-100) binding to CDK8 (1-403)/Cyclin C variants A 1-464 CycC 1-403 1-359 CycC B CycC MED12 1-100 1.0 [kDa] [kDa] [kDa] [kDa] CDK8 (1-464)/CycC, Kd: 044 nM 100 100 100 50 CDK8 (1-403)/CycC, K : 054 nM 75 75 d 75 37 CDK8 (1-359)/CycC, Kd: 220 nM 50 50 50 25 37 20 0.5 37 37 on Bound [-] 25 25 15
25 20 20 Frac 20 10 15 15 0.0 15 10 10 10 1.0E+00 1.0E+01 1.0E+02 1.0E+03 1.0E+04 1.0E+05 CDK8 CDK8 CDK8 CDK8 MED12 (1-100) concentra on [nM] C 1-403 CycC 1-403 CycC 1-403 CycC 1-403 CycC D 57.1 55.1 MED12 MED12 MED12 MED12 55 23-69 1-100 1-350 1-440 [kDa] [kDa] [kDa] [kDa] 50.6 100 100 100 100 75 75 75 75 50 Threshold 50 50 50 50 47.7 CDK8 (1-403)/CycC CDK8 (1-464)/CycC 37 37 37 37 [°C] Temperature MED12 (1-100)/CDK8 (1-403)/CycC 45 MED12 (1-100)/CDK8 (1-464)/CycC 25 25 25 25 20 20 20 20 55.4 55 54.6 15 15 15 15 52.3 10 10 10 10 50 49.4 Threshold CDK8 (1-403)/CycC MED12 (1-100)/CDK8 (1-464)/CycC CDK8 CDK8 [°C] Temperature MED12 (1-91)/CDK8 (1-403)/CycC E 1-403 CycC 1-403 CycC 45 MED12 (1-69)/CDK8 (1-403)/CycC MED12 CDK8 CDK8 CDK8 no MED12 1-100 F 1-403 CycC 1-403 CycC 1-403 CycC [kDa] [kDa] [min] -sub 2 5 -sub 2 5 -kin -kin 15 15 MED12 MED12 37 37 STAT1 no MED12MED12 1-100 MED12 11-69
[kDa] [min] 2 -sub -sub -kin 2 5 15 -sub 15 15 5 2 5 [kDa] [kDa] [min] -kin 5 2 -sub 15 -kin 5 15 -sub 2 37 STAT1 150 150 SIRT1 100 100 no MED12 MED12 1-100 MED12 23-69 [kDa] [kDa] [min] 2 -sub 5 -kin -sub 15 2 -kin 5 15 [kDa] [min] 5 5 -kin -sub 2 -sub 5 -sub 2 15 15 2 15 250 250 RPB1 37 STAT1
Fig. 1. MED12 both binds and activates binary CDK8/Cyclin C complexes. (A) SDS/PAGE analyses of purified binary CDK8/Cyclin C complexes and purified MED12 (1–100). Protein complexes are shown as cartoons above each gel, individual proteins are indicated next to the gels. CDK8 variants (1 to 464 [full- length], 1 to 403 and 1 to 359) are shown in red, Cyclin C is in blue, and MED12 is depicted in gray. (B) MST binding experiments using of MED12 (1–100) and
different binary CDK8/Cyclin C complexes. Kd values are indicated. Error bars reflect the SD of four replicates. Please note that Kd cannot be read off directly due to the experimental necessity to use high protein concentrations. (C) SDS/PAGE analyses of purified ternary CDK8 (1–403)/Cyclin C/MED12 complexes. Protein complexes comprising MED12 constructs of different sizes are shown as cartoons above each gel, individual proteins are indicated next to the gels. Colors as in A. We note that MED12 (23–69) carries a C-terminal Strep-tag and hence runs higher than expected. (D) Thermal stability of binary CDK8/Cyclin C
complexes and ternary CDK8/Cyclin C/MED12 complexes determined using nanoDSF. TM values are indicated. The SD of three experimental replicates is shown as error bars. (E) In vitro kinase assays using purified binary CDK8 (1–403)/Cyclin C and ternary CDK8 (1–403)/Cyclin C/MED12 (1–100) complexes. The 0.25 pmol kinase complex was incubated with 75 pmol GST-tagged STAT1 transactivation domain (TAD) in the presence of an excess of [γ-32P]-ATP. For SIRT1, a ratio of 0.5 pmol kinase and 50 pmol SIRT1 was used. For Pol II, 2 pmol of kinase complex was incubated in presence of 2 pmol Pol II. Please note that the Pol II C-terminal domain contains about 50 potential CDK8 phosphorylation sites, which made this 1:1 ratio necessary. Reactions were stopped at the indicated time points. (F) Same as in E; however, 7.5 pmol kinase complexes was incubated with 50 pmol STAT1 TAD.
2of12 | www.pnas.org/cgi/doi/10.1073/pnas.1917635117 Klatt et al. Downloaded by guest on September 27, 2021 and SI Appendix, Fig. S1A). As the N terminus of MED12 was at the MED12 N terminus (MED12 residues 23 to 69) (Fig. 1F reported to harbor the CDK8 activation domain (19), we mea- and SI Appendix, Fig. S1E). sured the affinity of MED12 (1–100) to purified CDK8/Cyclin C complexes using microscale thermophoresis (MST). Using MST, MED12 Makes Extensive Contacts to CDK8 without Substantially we found a nanomolar affinity (Kd = 44 nM) of MED12 for full- Contacting Cyclin C. Prior work on the mechanism of CDK8 ac- length CDK8 (Fig. 1B). Shortening of the CDK8 C terminus tivation by MED12 identified a surface groove on Cyclin C that reduced MED12 binding affinity by a factor of five for the was suggested to constitute the MED12 binding site necessary shortest CDK8 variant (CDK8 [1–359], Kd = 220 nM). This for CDK8 activation (19, 20, 22). However, how such a distant suggests that the disordered C terminus of CDK8 only margin- MED12 binding site would activate CDK8 is controversial. ally influences MED12 binding. To cross-validate our MST Therefore, to determine the location of MED12 in ternary measurements, we used isothermal titration calorimetry (ITC) to complexes, we carried out protein cross-linking coupled to mass confirm an affinity of 72 nM of MED12 (1–100) for CDK8 (1–403)/ spectrometry (XL-MS) (22). On the one hand, we utilized both – Cyclin C (the Kd using MST was 54 nM) (Fig. 1B and SI Appendix, 0.125 and 0.25 mM disuccinimidyl suberate (DSS) to form lysine Fig. S1B). Having determined a nanomolar affinity of MED12 for lysine cross-links in ternary CDK8 (1–403)/Cyclin C/MED12 binary CDK8/Cyclin C complexes, we aimed to develop expression complexes with varying lengths of MED12 (Fig. 2A). On the and purification strategies for ternary CDK8/Cyclin C/MED12 other hand, we obtained an independent set of distance re- complexes. Indeed, using MultiBacTurbo coexpression we were able straints for CDK8 (1–403)/Cyclin C/MED12 (1–100) from to purify a variety of ternary complexes that comprised MED12 carboxyl-specific and zero-length restraints introduced by the fragments with a length of up to 440 residues (Fig. 1C). Taking addition of 3.6 mg/mL pimelic acid dihydrazide (PDH) and CDK8 (1–403)/Cyclin C/MED12 (1–100) as a representative of the 4.8 mg/mL coupling reagent DMTMM (4-(4,6-Dimethoxy-1,3,5- ternary complexes investigated in this study, we validated its stoi- triazin-2-yl)-4-methylmorpholinium chloride) (Fig. 2B and SI chiometry using static light scattering, which revealed a mono- Appendix, Fig. S2C) (23). To properly assess the influence of disperse complex with a 1:1:1 stoichiometry (SI Appendix,Fig. MED12 binding on CDK8/Cyclin C, we also carried out DSS S1C). Furthermore, the minimal stable ternary complex could be XL-MS for binary CDK8/Cyclin C complexes (SI Appendix, formed with a MED12 fragment comprising residues 23 through 69 Fig. S2A). (Fig. 1C). To assess whether MED12 binding to different CDK8/ DSS-mediated cross-linking of CDK8 (1–403)/Cyclin C/MED12 Cyclin C complexes increases their stability, we measured their (1–100) resulted in the identification of 8 intersubunit cross-links thermal melting behavior using differential scanning fluorimetry and 45 intrasubunit cross-links using a DSS concentration of (nanoDSF). We found a significantly increased melting tempera- 0.125 mM (Fig. 2A and SI Appendix, Fig. S2B). A DSS concen- BIOCHEMISTRY ture (+7 °C) of ternary complexes, confirming that MED12 has a tration of 0.25 mM resulted in 6 intersubunit and 33 intrasubunit major stabilizing effect on CDK8/Cyclin C complexes (Fig. 1D). cross-links (Fig. 2A and Dataset S1). These numbers increased Note here that the stability of CDK complexes was reported to for ternary complexes encompassing MED12 (1–350) and MED12 correlate with their activity (20). Moreover, our nanoDSF data (1–440) (SI Appendix, Fig. S2B). The average length for all DSS uncover that the presence of CDK8’s likely unstructured C termi- cross-links is 15.8 Å, falling well into the reported range for DSS nus destabilizes both binary and ternary complexes. We find that cross-link distances (calculated using PDB ID code 5BNJ) (23, ternary complexes comprising a MED12 of fragments 1 to 91 or 1 24). The carboxyl-specific cross-linker was only successful for the to 100 result in the most stable complexes. In contrast, MED12 CDK8 (1–403)/Cyclin C/MED12 (1–100) complex. It allowed for constructs shorter than 70 residues significantly destabilize the the identification of one intersubunit acidic cross-link and four ternary complex, albeit still showing a stabilizing effect with respect intrasubunit acidic cross-links (SI Appendix, Fig. S2C and Dataset to the binary complex (Fig. 1D). S2). Moreover, DMTMM catalyzed the formation of 6 inter- MED12 binding to CDK8/Cyclin C was reported to activate subunit and 52 intrasubunit cross-links with zero-length distance the otherwise negligible kinase activity of CDK8 (19). We sought (Fig. 2B). These cross-links are extremely valuable due to their to recapitulate these findings using our purified binary CDK8/ high distance precision (23). Cyclin C and ternary CDK8/Cyclin C/MED12 complexes. In First and foremost, we detected 25 high-confidence cross-links addition, we asked which fragment of MED12 is sufficient to of MED12 to CDK8, but only 4 cross-links of MED12 to Cyclin activate CDK8. To measure CDK8 activity, we established a C (although Cyclin C contains many cross-linkable residues). radioactive kinase assay with the purified transactivation domain This strongly suggests that MED12 primarily contacts CDK8, not of STAT1 as a substrate (Fig. 1E). STAT1 is a known CDK8 Cyclin C. In addition, the number of intrasubunit cross-links is target and is phosphorylated by CDK8 at Ser-727 (17, 21). greatly diminished once MED12 binds to binary CDK8/Cyclin C Whereas we detected low kinase activity for binary CDK8/Cyclin complexes (Fig. 2A and SI Appendix, Fig. S2A). Next, we analyzed C complexes using our in vitro system, the inclusion of MED12 the distribution of all intersubunit cross-links involving MED12. in ternary CDK8 (1–403)/Cyclin C/MED12 (1–100) complexes This revealed three major sites of interaction between MED12 resulted in a pronounced stimulation of kinase activity (Fig. 1E and CDK8/Cyclin C (Fig. 2C). The first site comprises MED12 res- and SI Appendix, Fig. S1D). To extend our findings to more than idues 30 to 42, which interact with five residues in the C-lobe of a single CDK8 substrate, we also established Sirtuin 1 (SIRT1) CDK8 (K322, K314, K272, K271, K265). Interestingly, these five and Pol II as known CDK8 targets in our in vitro assay system. residues are located in an area of conserved protein–protein Using these substrates, we observed similar MED12-dependant interactions across the CDK family (18, 25). Moreover, we also activation of CDK8 (Fig. 1E). This confirms earlier reports (5, found this part of MED12 (30–42) to contact the αC helix of 19), yet also demonstrates that the CDK8/Cyclin C complex CDK8 (E66) and two acidic residues in Cyclin C (E98, E99). possesses basal kinase activity whose extent varies with the ratio Interestingly, E99 in Cyclin C was hypothesized to be involved in of kinase to substrate concentration (Fig. 1E and SI Appendix, CDK8 activation (18, 26). Taken together, these cross-links place Fig. S1D). Moreover, in addition to STAT1 phosphorylation (the residues 30 to 42 of MED12 between CDK8 and Cyclin C, in substrate), we also observed signal for CDK8 phosphorylation, proximity to the activation segment of the T-loop of CDK8 whose the extent of which varied dependent on the exact assay condi- structure was not resolved in the binary complex (18). The second tions and the used substrate (SI Appendix, Fig. S1D). Finally, site of MED12 interaction with CDK8 involves contacts between using the aforementioned assay, we systematically truncated the MED12 residues K60 and K68 and CDK8 residues K119 and MED12 N terminus. This allowed us to confine the minimal K303 (Fig. 2C). This contact site is located close to the C terminus of stable fragment sufficient for CDK8 activation to 46 amino acids CDK8, which was found to be disordered by X-ray crystallography
Klatt et al. PNAS Latest Articles | 3of12 Downloaded by guest on September 27, 2021 DSS DMTMM (Zero-length) A MED12 (1-100) MED12 B MED12 (1-100) CDK8 (1-403) 1 100 MED12 CDK8 (1-403) 1 100 Cyclin C Cyclin C
CycC 1 200 283 CycC 1 200 283
CDK8 1 200 403
CDK8 1 200 403
DSS MED12 (1-350) MED12 C 1 200 350 CDK8 (1-403) CycC Cyclin C 1 200 283 100 350 440
MED12 structured region CycC 1 200 400 1 200 283
CDK8 ebol-N ebol-C C-tail 1 200 403 CDK8 1 200 403
E98 E99 CycC K30
MED12 DSS Site I Site III MED12 (1-440) K30 K42 Site II K80 E141 MED12 K32 CDK8 (1-403) 1 200 400 440 E35 K93 K139 D34 K132 Cyclin C K60 K68 K99
CycC 1 200 283 E66 K74 K322 K265 K272 K307 K44 K271 K8 K314 CDK8 K303 K47 CDK8 1 200 403 C-lobe K119 N-lobe
disuccinimidyl suberate (DSS) pimelic acid dihydrazide (PDH/ZL)
K8 D Site III K30 K44 CDK8 N-lobeobe K47
K74 C-tail Site II E66 Cyclin C
KK115
E98/99 K307 K314 K303303K3 Site I K3222 MED12 CDK8DK8 K272/271K27K272/271 T-loop C-lobeC-lobee K26K265K265
Fig. 2. The N terminus of MED12 wraps around CDK8 and contacts its T-loop. (A) Schematic representation of all DSS-mediated cross-links obtained from ternary CDK8 (1–403)/Cyclin C/MED12 (1–100), CDK8 (1–403)/Cyclin C/MED12 (1–350), and CDK8 (1–403)/Cyclin C/MED12 (1–440) constructs using two different concentrations of DSS. Intramolecular cross-links are colored according to the individual protein subunits with CDK8 in red, Cyclin C in blue, and MED12 in gray. Intersubunit cross-links are shown as black lines. The endpoint of each line specifies a specific residue in the corresponding protein. All cross-links can be found in Dataset S1. Cross-linking data were displayed with help of xiVIEW (58). (B) Schematic representation of all cross-links detected after treatment of ternary CDK8 (1–403)/Cyclin C/MED12 (1–100) with PDH and DMTMM. Intramolecular cross-links are colored as in A. Intersubunit cross-links are shown as orange lines. All cross-links can be found in Dataset S2.(C, Upper) Cyclin C, MED12 (1–440) and CDK8 (1–403) with different protein domains highlighted. (Lower) A cartoon depiction of all intersubunit cross-links involving MED12. The intersubunit cross-links were grouped in three interaction sites (sites I to III) according to their localization on the surface of the binary CDK8/Cyclin C complex. (D) Three-dimensional arrangement of the three major sites of cross-links between MED12 and CDK8 (1–403)/Cyclin C. Lysine and glutamate residues that were found cross-linked are represented as black spheres and plotted onto the structure of the CDK8 (1–403)/Cyclin C complex (PDB ID code 3RGF) (18). Please note that K115 is shown instead of the structurally unresolved, cross-linked residue K119.
(18). In support of this contact, we note that we found slightly interaction site are situated on the surface of CDK8 that is ori- higher affinities of MED12 (1–100) for full-length CDK8 than ented away from its active site, explaining how MED12 can wrap for CDK8 (1–403) (Fig. 1B). This indicates a weak contribution around CDK8 without obstructing substrate binding. To exclude of the CDK8 C terminus to MED12 binding. A third site of that the C terminus of CDK8 alters the cross-linking patterns, we intersubunit cross-links involving MED12 encompasses MED12 repeated the XL-MS experiments using DSS chemistry with bi- residues K80 through E141. Except for K30 on Cyclin C, all of nary and ternary complexes comprising full-length CDK8 [CDK8 these residues establish cross-links to the N-lobe of CDK8. (1–464)] (SI Appendix,Fig.S2D). Our results revealed unchanged Interestingly, all MED12 cross-links that are part of the third cross-linking patterns, hence demonstrating that the CDK8 C
4of12 | www.pnas.org/cgi/doi/10.1073/pnas.1917635117 Klatt et al. Downloaded by guest on September 27, 2021 terminus does not impact the structure of the CDK8/Cyclin position. Much to our surprise, an E33Q mutation completely C/MED12 complex. Lack of cross-linkable residues in the CDK8 abolished CDK8 activation by MED12 (Fig. 3 A and B). This was C terminus did not allow to determine its exact position in the also the case for double-mutants involving E33. In contrast, complex. neither the mutation of D34 nor of E35 had an effect on kinase Taken together, our cross-linking data suggest a revised model activation. To exclude the direct involvement of other charged of how MED12 binds the binary CDK8/Cyclin C complex (Fig. residues in CDK8 activation, we measured the effect of K30A, 2D). The N-terminal portion of MED12 wraps around the entire Q31A, and K32A mutations on kinase activation. We found CDK8 molecule, making extensive contacts to both its C- and none of them to influence CDK8 activation (SI Appendix, Fig. N-lobe (Fig. 2 C and D). Most interestingly, our data allow us to S3A). We thus conclude that MED12 E33 is essential to activate place a stretch of about 10 amino acids (MED12 residues 30 to CDK8, prompting us to term the helix that harbors E33 at its tip 42) in immediate proximity of the T-loop of CDK8. This suggests “activation helix.” We note here that we were able to purify all that MED12 directly contacts the T-loop of CDK8, thereby re- ternary, mutation-containing complexes to homogeneity, dem- solving how MED12 is able to activate CDK8 in the absence of onstrating that MED12 binding to CDK8/Cyclin C and CDK8 T-loop phosphorylation. The second and third site of interaction activation can be experimentally separated. between MED12 and CDK8 illustrate that the N terminus of We next asked how the activation helix is able to bind in the MED12 folds around the entire CDK8 molecule (Fig. 2 C and interface of CDK8 and Cyclin C. To that end we calculated the D). K139 through E141 are the last residues of MED12 for which electrostatic surface potential of the binary CDK8/Cyclin C com- we detected cross-links to the CDK8/Cyclin C complex. Further- plex and of a model of the MED12 activation helix (SI Appendix, more, our cross-linking data of ternary complexes comprising Fig. S3B). We noticed a basic patch at the interface of CDK8 and MED12 constructs of various length suggest that the N terminus Cyclin C in the same region where we had detected cross-links of of MED12 is only structured until residue 174 (Fig. 2A). Beyond MED12 to CDK8. This warrants our speculation that the acidic this residue, we neither detect intra-MED12 cross-links nor cross- triad E33-D34-E35 is responsible for positioning the MED12 ac- links of MED12 to CDK8 or Cyclin C, hinting at the possibility tivation helix properly for CDK8 activation. that, beyond residue 174, MED12 is disordered in ternary com- plexes that lack MED13. Glutamate-33 Likely Positions the Arginine Triad of CDK8 and Thereby Activates the Kinase. Having discovered a helix in MED12 that is An Activation Helix in MED12 Is Essential for CDK8 Activation. Our crucial for its CDK8 activation potential, we next asked which
cross-linking data place MED12 residues 30 to 42 in vicinity of CDK8 residues contact E33 in MED12. Our quest was guided by BIOCHEMISTRY the T-loop of CDK8, suggesting that they are crucial for kinase a comparison of CDK8 with a CDK homolog, the phosphate- activation. Interestingly, secondary structure prediction clearly dependent signaling complex Pho85/Pho80. In this complex, a hinted at the formation of an α helix for MED12 residues 32 salt bridge between R132 on the C lobe of Pho85 (the CDK) and through 44. To assess the importance of this α-helix for CDK8 D136 on Pho80 (the Cyclin) locks the activation loop of Pho85, activation, we systematically mutated residues that are part of thereby circumventing the requirement for T-loop phosphory- the helix. In particular, we were intrigued by a cluster of three lation (27). A similar phosphorylation-independent mechanism acidic residues (E33, D34, and E35) at the predicted N-terminal was observed for CDK5/p25. There, p25 tethers the unphos- tip of the helix that stabilize the positive helical dipole at this phorylated T-loop of CDK5 in an active conformation (28).
no MED12 MED12 E33Q E33Q/D34N Side-View Top-View A [kDa] B 5 5 5 2 2 2
15 [min] 15 15 -sub -sub -sub -kin -kin E33 E33 STAT1 E35 CDK8 37 1-403 CycC WT MED12 D34N MED12 E35Q D34 E35 D34 MED12 [kDa] 5 5 5 2 2 2 [min] 15 15 15 -sub -sub -sub
-kin N-term. 1-100 -kin 37 STAT1 N-term.
WT CDK8 R65Q CDK8 WT 1-403 CycC [kDa] [min] 5 30 5 30 15 15 -sub -sub C -kin D 5 30 15 -sub STAT1 [kDa] -kin 37 MED12 STAT1 1-100 CDK8 CDK8 R150Q CDK8 R178Q 37 1-403 CycC CycC E98Q CycC E99Q E98Q/E99Q ]aDk[ ]nim[ 30 5 5 30 15 15 -sub -sub -kin 30 30 5 5 5 30 [min] 15 15 15 [kDa] -sub -sub -sub 37 STAT1 37 STAT1 WT CDK8 R65Q WT
[kDa] [min] [kDa] 2 15 -sub 5 30 30 5 30 5 15 15 -sub -sub -kin STAT1 STAT1 CDK8 37 37 1-403 CycC CDK8 R150Q CDK8 R178Q CycC N181A CycC D182A
]aDk[ ]nim[ 5 30 30 5 15 15 -sub -sub [kDa] -kin -sub 5 -sub 30 MED12 2 15 30 2 5 15 1-100 STAT1 37 37 STAT1
Fig. 3. MED12 utilizes an activation helix to stimulate the activity of CDK8. (A) In vitro kinase assays using ternary CDK8 (1–403)/Cyclin C/MED12 (1–100) complexes that harbor point-mutations within MED12. Kinase assays were carried out with the STAT1 TAD as described in Fig. 1F. WT denotes wild-type MED12. (B) Three-dimensional model of the MED12 activation helix comprising MED12 residues 19 to 50. The model was calculated using PEP-FOLD3 (59). The negatively charged triad of amino acids at the N-terminal tip of the activation helix (E33, D34, and E35) is shown as sticks in pink. (C) In vitro kinase assays using binary CDK8 (1–403)/Cyclin C and ternary CDK8 (1–403)/Cyclin C/MED12 (1–100) complexes. Each complex harbored a mutation in the CDK8 arginine triad. Kinase assays were carried out with the STAT1 TAD as a substrate as in Fig. 1F.(D) In vitro kinase assays using ternary CDK8 (1–403)/Cyclin C/MED12 (1– 100) complexes that harbor point-mutations within Cyclin C. The activity of wild-type ternary CDK8 (1–403)/Cyclin C/MED12 (1–100) is shown twice since the experiments were analyzed on separate gels. Kinase assays were carried out with the STAT1 TAD as a substrate as described in Fig. 1F.
Klatt et al. PNAS Latest Articles | 5of12 Downloaded by guest on September 27, 2021 Just like all human CDKs, CDK8 possesses an arginine triad mutations perfectly mirror the activation helix that we describe (R65, R150, and R178). However, as it lacks a phosphorylated to be essential for CDK8 activation by MED12 (Figs. 3B and T-loop residue that could coordinate these arginines, we asked 4C). As the mutational analysis of the activation helix uncovered whether one of the arginine residues instead contacts E33 in only a single amino acid to be critical for the CDK8-stimulatory MED12. To that end, we prepared individual arginine mutants function of MED12, we next examined the functional conse- (R65Q, R150Q, and R178Q) of CDK8 and tested those in quences of recurring disease mutations in MED12. First, we complex with Cyclin C and in ternary MED12-containing com- measured the affinity of MED12 (1–100) fragments carrying plexes (Fig. 3C). If one of the three arginine residues indeed con- L36R, Q43P, or G44S mutations for binary CDK8 (1–403)/ tacts E33 in MED12, we expected to see no effect of this mutation Cyclin C complexes. All three residues (L36, Q43, and G44) are on the basal kinase activity of the binary CDK8/Cyclin C complex. mutational hotspots in uterine leiomyomas and were also found In contrast, we predicted the abrogation of MED12-dependent in chronic lymphocytic leukemia (34, 37). In contrast to pre- CDK8 activation by such mutation. This is exactly what we de- viously published results, we did not detect significantly weak- C tected for all three arginine mutants (Fig. 3 ). However, we were ened binding of MED12 to binary complexes when compared to unable to detect significant and reproducible differences between wild-type MED12 (1–100) (Fig. 4A) (19). The same holds true the three arginine mutations, despite the fact that R65 is located in A α for MED12 carrying an E33 mutation (Fig. 4 ). This raised the the C helix of CDK8 and R150 and R178 are located in its T-loop. possibility that MED12 mutations that occur in different cancers To exclude that the individual arginine mutants impair MED12 might not influence MED12 binding to CDK8/Cyclin C, only its binding to CDK8/Cyclin C, we measured the affinity of MED12 activation. To test this, we purified ternary complexes containing (1–100) for binary CDK8 (1–403)/Cyclin C complexes carrying in- mutated MED12 variants (L36R, Q43P, and G44S) and mea- dividual arginine mutations (R65Q, R150Q, or R178Q) by MST sured their potential to activate the CDK8 kinase. As expected, without detecting major changes in the affinity of MED12 for for all complexes kinase activation was abolished or at least CDK8/Cyclin C (SI Appendix,Fig.S3C). drastically reduced (Fig. 4B). Interestingly, the nature of the In summary, our data establish that both the CDK8 arginine triad and E33 of MED12 are essential for MED12-dependent mutation has a profound impact on the activation potential of CDK8 activation. Whether the active conformation of the CDK8 MED12. Whereas the mutation of Asp-34 to Tyr (D34Y), which T-loop is induced by a direct salt bridge between E33 of MED12 is found in uterine leiomyomas, drastically reduces CDK8 kinase and one of the members of the CDK8 arginine triad will require activity, a D34N mutation has no impact on kinase activity in our A B high-resolution structural information on the ternary CDK8/ in vitro system (Figs. 3 and 4 ). Taken together, we find that Cyclin C/MED12 complex. mutations in MED12 that are found in cancer patients lead to an abrogation of CDK8 kinase activation without altering the af- Cyclin C Is Not Directly Involved in MED12-Dependent CDK8 Activation. finity of MED12 for CDK8/Cyclin C. Cyclin binding triggers the repositioning of the αC helix in cyclin- dependent kinases, thereby allowing for the formation of their Type II Kinase Inhibitors Show Reduced Efficacy toward Ternary active site (29, 30). This mechanism also applies to binding of CDK8/Cyclin C/MED12 Complexes. Due to its prominent oncogenic Cyclin C to CDK8 (18). In addition to this conserved role, res- role, considerable efforts were made to develop CDK8-specific idue E99 in Cyclin C and an unusually deep surface groove be- inhibitors (17, 21, 37, 38). As we discovered that MED12 tween its two cyclin boxes were also suggested to contribute to CDK8 activation (16, 17, 23, 26). We sought to confirm these results by using our in vitro kinase assay system. However, to our MED12 (1-100) mutants binding to CDK8 (1-403)/Cyclin C Side-View A MED12 1-100 WT, K : 54 nM C surprise neither an E98Q or E99Q mutation, nor the mutation of d L36 Q43 1.0 MED12 1-100 G44S, Kd: 41 nM two surface groove residues (N181A, D182A) on Cyclin C that were G44 MED12 1-100 Q43P, K : 34 nM suggested to bind MED12 (19, 31) had an impact on MED12- d MED12 1-100 L36R, K : 11 nM D d driven CDK8 activation (Fig. 3 ). Using nanoDSF measure- MED12 1-100 E33Q, K : 31 nM D34 ments, we are able to show that both Cyclin C surface mutants 0.5 d (N181A, D182A) destabilize ternary MED12-containing complexes N-term. SI Appendix D ( ,Fig.S3 ). Taken together, our functional and struc- 0.0 tural data suggest that the Cyclin C surface groove is not directly involved in the MED12-dependent activation of CDK8. Rather, Top-View
upon mutation of residues in the surface groove, the entire complex no MED12WT MED12 L36R B L36 G44 becomes destabilized and thereby leads to reduced kinase activity [kDa] [min] 30 5 30 5 5 -sub 15 -sub 30 15 -sub -kin 15 Q43 STAT1 under specific experimental conditions (19). Alternatively, the sur- CDK8 37 1-403 CycC face groove might contribute to substrate recognition or serve as a D34 MED12 MED12 Q43P MED12 G44S MED12 D34Y N-term. protein interaction site, as is the case for CDK2/Cyclin A (32, 33). 1-100 [kDa] [min] 5 30 5 5 30 30 15 15 -sub -sub -sub -kin In conclusion and despite the fact that we detected cross-links for 15 STAT1 Cyclin C E98 and E99 to the MED12 activation helix (Fig. 2C), 37 Cyclin C seems only of minor importance for the MED12-driven Fig. 4. Cancer-associated mutations within the MED12 activation helix activation of CDK8. abolish CDK8 activation without altering MED12 affinity for CDK8/Cyclin C. (A) MST binding experiments using MED12 (1–100) variants carrying different Malignant MED12 Mutations Retain Their Affinity for CDK8, yet Abolish cancer-associated mutations and binary CDK8/Cyclin C complexes. MED12 – MED12-Dependent CDK8 Activation. Frequent somatic MED12 mu- (1 100) E33Q serves as activation-dead control (Fig. 3A). Kd values are in- tations are associated with, among others, uterine leiomyomas, dicated. Error bars reflect the SD of four replicates. Please note that Kd tumors of the breast, and chronic lymphocytic leukemia (34–37). cannot be read off directly due to the experimental necessity to use high- protein concentrations. (B) In vitro kinase assays using ternary CDK8 (1–403)/ Despite the fact that MED12 consists of 45 exons, the over- – whelming majority of these mutations are found in MED12 exon Cyclin C/MED12 (1 100) complexes harboring individual cancer-associated mutations within MED12. Kinase assays were carried out with the STAT1 2 (amino acids 34 to 68) and to a lesser extent in exon 1 (amino TAD as a substrate as in Fig. 1F.(C) Three-dimensional model of the MED12 acids 1 to 33) (34). Strikingly, when we mapped the mutations activation helix comprising MED12 residues 19 to 50. The model was calcu- occurring in chronic lymphocytic leukemia (34) onto the se- lated using PEP-FOLD3 (59). MED12 residues frequently mutated in cancer quence of MED12 (Fig. 2D), we found that the patient-derived are shown as violet sticks (D34Y, L36R, Q43P, and G44S).
6of12 | www.pnas.org/cgi/doi/10.1073/pnas.1917635117 Klatt et al. Downloaded by guest on September 27, 2021 employs an activation helix to stimulate CDK8 activity, we asked the binary and the ternary CDK8 complex (Fig. 5A and SI Ap- whether the presence of MED12 alters the efficacy of commer- pendix, Fig. S4A). However, strikingly, all type II inhibitors cially available inhibitors, all of which were developed against the showed a significantly diminished affinity for the ternary CDK8 binary CDK8/Cyclin C complex (38). We suspected that this complex. Only sorafenib (a type II inhibitor) displaced the re- might be the case because we found MED12 to contact the αC porter probe by more than 50%, albeit with an affinity decreased helix of CDK8 (Fig. 2). Moreover, MED12 likely positions the by two orders-of-magnitude (Fig. 5A and SI Appendix, Fig. S4A). CDK8 T-loop in a catalytically competent conformation by These results indicate that type II kinase inhibitors, which bind contacting one or several members of its arginine triad (Fig. 3). to the CDK8 hinge region and extend to the so-called deep In further support of a differential efficacy of commercial in- pocket (18), are hindered from binding to CDK8 when MED12 hibitors against MED12-bound CDK8 complexes, we note that is present. The deep pocket is only accessible when CDK8 is in type II inhibitors, such as sorafenib, were previously reported not its inactive state (DMG-out conformation of CDK8), suggesting to translate their inhibitory efficacy into the cellular context (21). that MED12 binding to CDK8 induces a DMG-in–like confor- To test our hypothesis, we selected three type I kinase inhib- mation that renders type I kinase inhibitors more potent once itors: Compound A (PDB ID code 6T41) (SI Appendix, Fig. S4), CDK8 is part of the entire Mediator kinase module. CCT251545 (PDB ID code 5BNJ), and BI-1347 (BI-1347 was To validate this conclusion, we next used MST to measure the chosen as a type I inhibitor due to its chemical similarity to affinity of the MED12 (1–100) fragment for binary CDK8 CCT251545 and its high in vivo potency [opnMe Boehringer (1–403)/Cyclin C complexes in presence of saturating levels of a Ingelheim Open Innovation Portal]). Moreover, we chose three type I (CCT251545) or a type II inhibitor (sorafenib). The af- known type II inhibitors: Sorafenib (PDB ID code 3RGF), finity of MED12 (1–100) for CDK8 (1–403)/Cyclin C prebound compound 2 (PDB ID code 4F7L), and BIRB796 (39), and to CCT251545 was about 21 nM, just like for the apo CDK8 measured the dissociation constants and IC50 values of both (1–403)/Cyclin C control. In contrast, a binary complex prebound classes of inhibitors toward a binary CDK8 (1–403)/Cyclin C and to the type II inhibitor sorafenib showed a fivefold reduced af- a ternary CDK8 (1–403)/Cyclin C/MED12 (1–100) complex (Fig. finity for MED12 with a Kd of about 130 nM (Fig. 5B). In further 5A and SI Appendix, Fig. S4A). To that end we used a reporter support of these results, we used ITC to, yet again, find a fivefold displacement assay that we had developed for CDK8/Cyclin C reduction in the affinity of MED12 (1–100) for CDK8 (1–403)/ previously (40). In analogy to the reporter probe (SI Appendix, Cyclin C prebound to sorafenib as compared to prebinding of Fig. S4B), all type I inhibitors displayed similar affinities for both CCT251545, which only showed a minor reduction of MED12 BIOCHEMISTRY
1.0E+04 MED12 (1-100) binding to CDK8 (1-403)/Cyclin C CDK8 (1-464)/CycC A Reporter B pre-bound to CCT254515 or sorafenib MED12 (1-1232)/CDK8 (1-464)/CycC Displacement 1.0E+03 <50%
1.0 CDK8 (1-403)/CycC + CCT254525 K : 021 nM 1.0E+02 d CDK8 (1-403)/CycC Kd: 023 nM
CDK8 (1-403)/CycC + Sorafenib, Kd: 128 nM 1.0E+01 0.5
1.0E+00 [µM] d K
0.0 1.0E-01
1.0E-02
1.0E-02 1.0E-01 1.0E+00 1.0E+01 1.0E+02 1.0E+03 1.0E+04 1.0E-03
No detectable Type-I Type-II D CCT254515 BI-1347 Compound A Compound 2 BIRB796 Sorafenib 1.0E+04 CDK8 (1-464)/CycC MED12 (11-91)/CDK8 (1-403)/CycC C MED12 (1-100) binding to CDK8 (1-403)/Cyclin C MED12 (1-1232)/CDK8 (1-464)/CycC pre-bound to CCT254515 or sorafenib 1.0E+03 Time (s) 0 1000 2000 3000 4000 5000 6000 7000 8000 1.0E+02
0
-0,1 1.0E+01 [µM]
-0,2 50 IC 1.0E+00 -0,3
-0,4 1.0E-01
20
0
-20 1.0E-02
-40 -60 CCT K : 72 nM d 1.0E-03 -80 Sorafenib K : 150 nM d -100 No inhibitor K : 29 nM -120 d
0,5 1,0 1,5 2,0 Type-I Type-II CCT254515 BI-1347 Compound A Compound 2 BIRB796 Sorafenib
Fig. 5. Type II kinase inhibitors lose part of their inhibitory potential when used against ternary CDK8/Cyclin C/MED12 complexes. (A) Reporter displacement assay (RDA) with three type I kinase inhibitors (CCT251545, BI-1347, and compound A) and three type II inhibitors (compound 2, BIRB 976, and sorafenib).
RDAs were measured using the full-length binary CDK8/Cyclin C complex and a ternary CDK8/Cyclin C/MED12 (1–1232) complex. Kd values of individual compounds were calculated using the Cheng–Prussoff equation as described previously (40). For clarity, Kd values of compound showing a probe displacement of less than 50% were set to 10 mM. (B) MST binding experiments using MED12 (1–100) and 2 nM binary CDK8 (1–403)/Cyclin C complexes that were either
preloaded with 30 μM of the type I inhibitor CCT251545 or with 30 μM of the type II inhibitor sorafenib. Apo CDK8 (1–403)/Cyclin C was used as a control. Kd values are indicated. Error bars reflect the SD of two replicates. (C) Using ITC, the MST measurements from (B) were confirmed, Kd values are indicated. (Upper) The corrected heat rates; (Lower) The calculated enthalpies per injection are plotted against the molar ratio of ligand and target. Please note that
CCT251545 addition resulted in some protein precipitate, which altered the observed molar ratio. (D)IC50 values for a binary CDK8 (1–464)/Cyclin C complex and two ternary complex variants [CDK8 (1–403)/CyclinC/MED12 (11–91), and CDK8 (1–464)/Cyclin C/MED12 (1–1232)] using the same set of inhibitors as in A
and ADP-Glo data. IC50 values of individual compounds were calculated using a standard XLfit algorithm (SI Appendix, Fig. S6). For clarity, IC50 values of compounds showing an enzymatic inhibition of less than 50% were set to 10 mM.
Klatt et al. PNAS Latest Articles | 7of12 Downloaded by guest on September 27, 2021 affinity for the binary CDK8 complex (Fig. 5C). We note here strand-specific libraries from HCT116 wild-type and E33Q cells in that a fivefold reduction in MED12 affinity for the binary CDK8 untreated condition and after 24 h of IFN-γ stimulation (Fig. 6B) complex is highly significant, as we do not expect the inhibitors to (45). The subsequent transcriptome-wide gene-expression analysis interfere with MED12 interaction sites II+III (Fig. 2D). Hence, uncovered the up-regulation of IFN response genes in both wild- we expected type II inhibitor prebinding to only reduce MED12 type and E33Q cell lines upon IFN-γ stimulation (Fig. 6B). affinity for CDK8 (1–403)/Cyclin C, which is precisely what we Moreover, when comparing E33Q cells to wild-type we found observed (Fig. 5 B and C). additional biological processes, such as tissue morphogenesis and Finally, we sought to corroborate our binding data by com- innate immunity, to be affected by IFN-γ treatment. In support of paring the enzymatic activity of binary full-length CDK8 (1–464)/ this finding, >40% of all IFN-γ–responsive genes are known to be Cyclin C with a minimal [CDK8 (1–403)/Cyclin C/MED12 (11–91)] regulated by CDK8-mediated STAT1 phosphorylation, thus and a longer ternary complex variant [CDK8 (1–464)/Cyclin identifying CDK8 as an important regulator of antiviral responses C/MED12 (1–1232)]. To achieve this, we utilized the ADP-Glo (46). Furthermore, our data confirm that IFN-γ stimulation results technology (Promega) to measure CDK8 activity. We limited our in a significant down-regulation of translation and translation- measurements to CDK8 phosphorylation rates here, as the in- related processes (Fig. 6B)(47). clusion of substrates rendered the ADP-Glo data hard to interpret Subsequently, we analyzed genes that were differentially due to the presence of multiple CDK8 substrates. This strategy is expressed between wild-type and E33Q mutant cells (Dataset S3). valid as we had previously shown that CDK8 serves as an alter- We identified 830 such genes; 131 genes were significantly up- native, second substrate in kinase assays (SI Appendix,Fig.S1D). In regulated (>log2 fold), whereas we found 62 to be down-regulated analogy to the reporter displacement assay, the ADP-Glo data ( 8of12 | www.pnas.org/cgi/doi/10.1073/pnas.1917635117 Klatt et al. Downloaded by guest on September 27, 2021 Downloaded by guest on September 27, 2021 eta hwn oeuae u-euae n onrgltd Otrsi idtp C16(T n 3Qkokn(I C16cls2 fe IFN- after h 24 cells HCT116 (KI) knockin E33Q and (WT) HCT116 wild-type in GO-terms ( down-regulated) treatment. and (up-regulated coregulated showing Heatmap htaeu-euae nMD2E3 nci C16cls 2i olcino uae eest nMiD.C ersnsocgncsgaue fcellular of signatures oncogenic represents C6 MSigDB. in sets gene curated of collection a is MSigDB. C2 cells. of HCT116 part knockin as E33Q pathways MED12 in up-regulated are that E1 3Qmtn el.Tedse oiotlrdln ersnsa represents line red horizontal dashed log The cells. mutant E33Q MED12 oklwt eeaea C16cl iecryn E1 3Qmtto.Tepeec fMD2i h E1 3Qmtn elln a eiidby verified was line cell mutant E33Q MED12 IFN- the upon in MED12 domain of transactivation presence STAT1 The the mutation. in E33Q Ser-727 MED12 of a phosphorylation carrying line The cell blotting. HCT116 Western an generate to workflow 6. Fig. lt tal. et Klatt 2 A E C sgRNA1 fold-changes log10p−value SMID_BREAST_CANCER_NORMAL_LIKE_UP 0.0084457 WALLACE_PROSTATE_CANCER_RACE_UP 0.0084285 LEF1_UP.V1_UP 0.0365194 10 20 30 40 0 oso E1-eedn D8atvto eebe rncitoa rflso ua acr.( cancers. human of profiles transcriptional resembles activation CDK8 MED12-dependent of Loss Cas9-2A -GFP 1 50510 5 0 −5 −10 Screening 62 genes C ( HCT116 down NGS) ocn ltsoiglog showing plot Volcano ) (transfec 1. delivery 4. propaga 5. library prepara (FACS 2. sor 3. single-cellisola ssODN > ) log 2or Gene setq-valueSetsizeMSigDB ng sgRNA2 2 on fold change on < ) − .( 2. on on D Otr nihetaayi fu-( up- of analysis enrichment term GO ) 131 genes 04824 Exon1: aa1-33 up WT 24 ssODN flanking 2 odcagdu-addw-euae ee 4hatrIFN-γ after h 24 genes down-regulated and up- changed fold sgRNAs target region for HDR; g.97G> [kDa] SAM−dependentmethyltransferase activity 250 D hydrolase activity, actingoncarbon−nitrogen(but notpeptide)bonds, inlinearamidine 50 048 E33Q E33Q (30nt) WT C E33Q Exon2: aa34-68 α-Tubulin STAT1 pS727 IFN-γ [hpt] MED12 transforming growth factor betareceptor, cytoplasmicmediatoractivity α-Tubulin C2 C2 C6 > P log au hehl ( threshold value 2 B histone methyltransferase activity(H3−K9specific) od n onrgltd( down-regulated and fold) log transforming growth factor betareceptorbinding Downregulated Upregulated histone−lysine N−methyltransferase activity protein−lysine N−methyltransferase activity p−value WT 0h vs WT 24h extracellular constituent structural matrix voltage−gated sodiumchannelactivity promoter−specific chromatinbinding E33Q 0h vs E33Q 24h 10 catalytic activity, actingonaprotein serine−type endopeptidaseactivity serine−type protein methyltransferase activity C2H2 zincfingerdomainbinding P GO:0045596 negative regulationofcelldifferentiation GO:0022604 regulationofcellmorphogenesis GO:0032318 regulationofRasGTPase activity GO:0035023 regulationofRhoproteinsignaltransduction GO:0045165 cellfate commitment GO:0071345 cellularresponsetocytokinestimulus GO:0009615 responsetovirus GO:0002252 immune effector process ofanepithelium GO:0002009 morphogenesis GO:0019221 cytokine−mediatedsignalingpathway GO:0030029 actinfilament−basedprocess GO:0060333 interferon−gamma−mediated signalingpathway GO:0031347 regulationofdefense response GO:0030036 actincytoskeleton organization GO:0046578 regulationofRasproteinsignaltransduction GO:0051607 defense responsetovirus GO:0051056 regulationofsmallGTPase mediatedsignaltransduction GO:0048729 tissuemorphogenesis GO:0071357 cellularresponsetotypeIinterferon GO:0060337 typeIinterferon−mediated signalingpathway GO:0034340 responsetotypeIinterferon GO:0006397 mRNAprocessing GO:0032984 macromolecularcomplex disassembly GO:0043241 proteincomplex disassembly GO:0043624 cellularproteincomplex disassembly GO:0008380 RNAsplicing NADHtoubiquinone electrontransport, GO:0006120 mitochondrial ATPGO:0042775 mitochondrial synthesiscoupledelectrontransport GO:0042773 ATP synthesiscoupledelectrontransport GO:0065004 protein−DNAcomplex assembly GO:0006334 nucleosome assembly GO:0006401 RNAcatabolicprocess GO:0006402 mRNAcatabolicprocess GO:0006612 proteintargetingtomembrane GO:0045333 cellularrespiration GO:0000956 nuclear−transcribed mRNAcatabolicprocess GO:0006412 translation GO:0000184 nuclear−transcribed mRNAcatabolicprocess, nonsense−mediateddecay GO:0019083 viral transcription GO:0019080 viral genomeexpression chain GO:0022900 electrontransport GO:0072594 establishment ofproteinlocalizationtoorganelle GO:0006414 translational elongation chain electrontransport GO:0022904 respiratory GO:0006413 translational initiation GO:0006415 translational termination GO:0070972 proteinlocalizationtoendoplasmicreticulum GO:0045047 proteintargetingtoER GO:0006614 SRP−dependentcotranslational proteintargetingtomembrane GO:0072599 establishment ofproteinlocalizationtoendoplasmicreticulum GO:0006613 cotranslational proteintargetingtomembrane < serine−type peptidaseactivity serine−type N−methyltransferase activity .5.Tetodse etcllnsilsrt h hehl of threshold the illustrate lines vertical dashed two The 0.05). signaling receptorbinding cytokine receptorbinding serine hydrolaseserine activity chemorepellent activity receptor ligandactivity 45678 γ arylsulfatase activity arylsulfatase calcium ionbinding nuto a loaaye yWsenbotn.( blotting. Western by analyzed also was induction chromatin binding R−SMAD binding A < ramn.Wl-yeHT1 el r oprdto compared are cells HCT116 Wild-type treatment. protein binding SMAD binding activin binding log cation binding ceai ersnaino h mlydCRISPR employed the of representation Schematic ) p53 binding 2 od ee.( genes. fold) Downregulated ( BIOCHEMISTRY against both binary and ternary complexes is likely due to its CDK8 N-lobe discovery in cell-based SILAC (stable isotope labeling by/with T-loop Step 1 CycC amino acids in cell culture) pulldown assays (21). There, CDK8 + Cyclin C is predominantly present in complex with MED12, which en- αC αC abled the discovery of CCT251545 as potent inhibitor of ternary CDK8 “pushed-out” CDK8 C-lobe “pushed-in” CDK8 complexes. Lastly, novel drugs intended at reducing inactive DMG-out basal activity + MED12 CDK8 activity may also be developed by using small molecules Step 2 that disrupt the interface between the MED12 activation helix Site III CDK8 and CDK8 (53). Such an approach would likely not impair the active MED12 N-term. structural integrity of the Mediator kinase module and may only CDK8 segment CDK8 specifically target MED12’s role in activating CDK8. substrateate CycC Site II CycC Our data demonstrate that reducing CDK8 activity can have R R both positive and negative effects on target gene expression (Fig. R E33 Site I “activationac helix” 6). This is likely the result of the multitude of CDK8 targets and tivati αC the differential functional outcome of target phosphorylation by on T-loop “pushed-in” CDK8 (11). For future drug development, it is therefore im- positioning DMG-in perative to first decipher the exact role of CDK8 in the particular cancer under investigation and then to decide whether reducing Fig. 7. Revised model of how MED12 activates CDK8. Step 1: Cyclin C binds to CDK8 and pushes the αC-helix of CDK8 into the “pushed-in” conforma- CDK8 activity is the right route to take to impact the disease. tion. This binding event is crucial for the formation of the active site of CDK8 This caution is warranted since MED12 was found to be involved and results in basal kinase activity as demonstrated in Fig. 1E. Step 2: MED12 in chemotherapy resistance without this finding having anything binding to CDK8/Cyclin C stabilizes and activates the entire ternary complex. to do with its nuclear function as a CDK8 activator and part of In particular, an activation helix in MED12 contacts and stabilizes the T-loop the Mediator kinase module (54). of CDK8, thereby activating the kinase. Likely, this contact is established through an interaction of an acidic residue at the N-terminal tip of the Materials and Methods MED12 activation helix (Glu-33) and the CDK8 arginine triad (Arg-65, Arg- Expression and Purification of Binary CDK8/Cyclin C and Ternary CDK8/Cyclin 150, and Arg-178). Moreover, MED12 binding favors the active site of CDK8 C/MED12 Complexes. To obtain binary CDK8/Cyclin C complexes, full-length to adopt a DMG-in conformation to enhance its activity. The preference Cyclin C (1–283) and different CDK8 variants (e.g., full-length [1 to 464] for a DMG-in conformation of the active site disfavors type II kinase inhib- and C-terminally truncated CDK8 [1 to 403, 1 to 359]) were cloned into itors from binding and inhibiting CDK8 in ternary CDK8/Cyclin C/MED12 pAceBac1 and pUCDM vectors, respectively. To enable affinity purification, complexes. Cyclin C carried a C-terminal Strep-tag and a tobacco etch virus (TEV) cleavage site. Ternary complexes were constructed similarly. However, they contained the C-terminal Strep-tag on multiple N- or C-terminally truncated three members of the CDK8 arginine triad. Since E66 is pointing MED12 variants. Both binary CDK8/Cyclin C and ternary CDK8/Cyclin inside the CDK8 molecule in binary CDK8/Cyclin C complexes C/MED12 complexes were coexpressed in High5 insect cells by recombinant this suggest a further rearrangement of the αC-helix upon baculovirus infection via a titerless protocol. For large-scale production of MED12 binding to CDK8 (SI Appendix, Fig. S8). Finally, we desired complexes, 4 to 6 L High5 cells at a density of 1.25 × 106 cells/mL in hypothesize that the contact between the arginine triad and the serum-free SF-4 Baculo Express media (Bioconcept) were infected in Fernbach activation helix of MED12 leads to a stably “folded-away” con- flasks. Flasks were incubated at 27 °C for 48 to 72 h on a rotary shaker formation of the CDK8 T-loop, allowing unobstructed substrate before cells were collected by centrifugation and resuspended in buffer A (50 mM Hepes/NaOH [pH 7.4], 150 mM NaCl, 2 mM DTT) supplemented with binding to the active site. protease inhibitors. Cells were lyzed by sonication and cell debris was re- A low-resolution cryoelectron microscopy map of the entire moved by centrifugation at 35,000 rpm for 90 min in a Beckmann Ti-45 ro- kinase module from Saccharomyces cerevisiae (52) was inter- tor. Supernatants were collected and bound to StrepTactin Superflow (IBA preted to show that MED12 contacts Cyclin C only, while our Lifesciences) resin followed by their elution with buffer A supplemented cross-linking, functional, and mutational data indicate the es- with 2.5 mM desthiobiotin. Cleavage of Strep-Tag fusions was performed sential role of the MED12 N-terminal segment and the CDK8 overnight with TEV-protease. After cleavage, tag-free protein complexes αC-helix for contact formation. This apparent discrepancy may were subjected to an anion-exchange column (ResourceQ, GE Healthcare). The flow-through after anion-exchange chromatography was subsequently be due to the limited resolution (15 Å) of the cryoelectron mi- loaded onto a cation-exchange column (Resource S, GE Healthcare) to re- croscopy map that cannot resolve the extended shape of the cover the desired protein complexes. In a final step the protein complexes MED12 N-terminal segment of and its detailed interactions with were polished by size-exclusion chromatography (SEC) using a Superdex 200 the kinase. 10/300 GL (GE Healthcare) or Superose6 10/300 GL (GE Healthcare) column in buffer B (50 mM Hepes/NaOH [pH 7.4], 100 mM NaCl, 2 mM DTT). Effective CDK8 Inhibitors Need to Be Developed Against MED12- Bound CDK8 Complexes. Whereas class I kinase inhibitors inhibit Chemical Cross-Linking and Sample Processing for Mass Spectrometry. To op- both binary CDK8/Cyclin C and ternary CDK8/Cyclin C/MED12 timize the cross-linking reaction conditions, pilot experiments were first μ complexes equally well, class II inhibitors lose a significant carried out at small scale (5 g protein) and monitored by SDS/PAGE. We fraction of their inhibitory potential when utilized against the selected conditions that resulted in an intensity reduction of the bands corresponding to monomeric subunits while at the same time preventing ternary CDK8/Cyclin C/MED12 complex (Fig. 5). This finding the formation of very high mass products. For the cross-linking reagent DSS, highlights a common problem in drug development: Whereas we chose two concentrations (125 and 250 μM). For the combination of PDH atomic resolution structures are needed for rational structure- and the activating reagent DMTMM, we selected concentrations of 3.6 and based drug design, most drug targets are part of larger com- 4.8 mg/mL, respectively. Final experiments for mass spectrometry were plexes, whose structures are unknown. Our study suggests that performed at a protein concentration of 1 mg/mL in cross-linking buffer future drug development aimed at CDK8-specific drugs needs to (20 mM Hepes pH 7.4, 100 mM NaCl, 2 mM DTT/0.5 mM TCEP) and with 50 μg focus on ternary, MED12-containing complexes. This is the case total protein. All cross-linking steps were carried out at 37 °C with mild — — shaking. DSS (a 1:1 mixture of d0- and d12-labeled forms, Creative Molecules) as our data suggest that MED12 induces or at least favors a was added as a freshly prepared 25 mM stock solution in anhydrous di- DFG-in conformation of the active site of CDK8 (Fig. 7), methyl formamide. After incubation for 30 min, the reaction was stopped by thereby essentially precluding efficient binding of class II inhib- adding ammonium bicarbonate to 50 mM. PDH (d0/d10) and DMTMM (both itors. The fact that we found CCT251545 to be equally effective from Sigma-Aldrich) were added from freshly prepared stock solutions in 10 of 12 | www.pnas.org/cgi/doi/10.1073/pnas.1917635117 Klatt et al. Downloaded by guest on September 27, 2021 sample buffer. After incubation for 45 min, the reaction was stopped by system. The ssODN contained phosphorothioate modifications and encom- passing the sample solution through Zeba spin desalting columns (7 K mo- passed the nucleotide to be mutated (g.97 G→C) (56). It was designed using lecular weight cutoff, ThermoFisher Scientific). For both chemistries, cross- the CRISPR Design Tool from Dharmacon. HCT116 cells were cotransfected in linked samples were evaporated to dryness in a vacuum centrifuge. Further six-well plates at 80 to 90% confluency with either PX458-MED12-gRNA#1 or sample processing followed standard procedures for reduction of disulfide PX458-MED12-gRNA#2 plasmids and 5 μL ssODN (100 μM) using Lipofect- bonds (with TCEP), alkylation of free thiols (with iodoacetamide), and en- amine 2000 according to the manufacturer’s instructions. Five hours post- zymatic digestion with endoproteinase Lys-C (Wako, 1:100, 37 °C for 2 to 3 h) transfection, cells were transfected again with 5 μL ssODN (100 μM) using and trypsin (Promega, 1:50, 37 °C overnight). Digests were purified by solid- phase extraction using SepPak tC18 cartridges (Waters), and purified pep- Lipofectamine 2000. Forty-eight hours posttransfection, positive GFP- tide mixtures were enriched for cross-linked peptides using SEC fractionation expressing cells were collected by using a FACSAria II cell sorter (BD Biosci- on a Superdex Peptide PC 3.2/30 column (GE Healthcare), as described ences). Single cell-derived clones were subsequently cultivated and positive previously (55). Three to four SEC fractions were collected for LC-MS/MS clones were confirmed by sequencing. analysis. MED12 gRNA#1: AAGGAGGTGCGTTCGAAAAT Liquid Chromatography-Tandem Mass Spectrometry. LC-MS/MS was per- MED12 gRNA#2: CGAACGCACCTCCTTCTGTT formed on an Orbitrap Fusion Lumos mass spectrometer coupled to an Easy ssODN sequence (*: phosphorothioates): nLC-1200 HPLC system via a Nanoflex electrospray source (all ThermoFisher Scientific). Peptides were separated by reversed-phase chromatography on an G*A*TGTTTACCCTCAGGACCCCAAACAGAAGCAGGTGCGTTC- Acclaim PepMap RSLC C18 column (250 mm × 75 μm, ThermoFisher Scien- GAAAATCGGGGCTCTGGAG*G* tific). Gradient elution was performed with the mobile phase A = water/ acetonitrile/formic acid (98:2:0.15 [vol/vol/v]) and B = acetonitrile/water/ formic acid (80:20:0.15 [vol/vol/v]) and a gradient from 11 to 40%B in 60 min Data Availability. The mass spectrometry data were deposited to the Pro- at a flow rate of 300 nL/min. The mass spectrometer was operated in data- teomeXchange Consortium via the PRIDE partner repository (57) (dataset dependent acquisition mode using top-speed mode with a cycle time of 3 s. identifier PXD015394). The structure of Compound A in complex with CDK8 Full-scan mass spectra were acquired in the Orbitrap analyzer at 120,000 (1–403)/Cyclin C were deposited to the Protein Data Bank (PDB ID code nominal resolution. Precursors with a charge state of +3to+7 were selected 6T41). All sequencing data were deposited in the Gene Expression Omnibus for fragmentation with quadrupole isolation and an isolation window of 2.0 (GEO), series GSE135458. m/z. Fragmentation was performed using collision-induced dissociation in the linear ion trap and detected in the linear ion trap in rapid scan mode. ACKNOWLEDGMENTS. We thank Silke Spudeit, Sooruban Shanmugaratnam, Dynamic exclusion was activated for 30 s after one scan event. Marc Heinrichmeyer, and Kathrin Oertwig for technical help; Norbert Eichner for performing sequencing analysis on CRISPR clones; Martin Humenik In Vitro Kinase Assays. All in vitro kinase assays were carried out with purified for static light-scattering experiments; Ramona Adolph and Dr. Clemens BIOCHEMISTRY binary or ternary complexes. Reactions contained 1 to 5 pmol of CDK8 and 50 Steegborn for providing the SIRT1 expression plasmid; and Dr. Lars Neumann to 200 pmol of substrate in kinase buffer (25 mM Tris/HCl [pH 8.0], 100 mM and Dr. Klaus Maskos from Proteros Biostructures GmbH for help with KCl, 10 mM MgCl2, 0.1 mM EGTA, and 2 mM DTT). Kinase reactions were biochemical and biophysical experiments and their analysis. This work was 32 started by the addition of 1 μCi γ-[ P]ATP and subsequent incubation at supported by the Elite Network of Bavaria, the University of Bayreuth, and the 37 °C. Reactions were separated by SDS/PAGE. Gels were dried and exposed Paul Ehrlich and Ludwig Darmstaedter Prize for Young Researchers (to C.-D.K.); to a phosphoimaging plate for 0.5 to 2 h. Images were recorded with a CR 35 and the Deutsche Forschungsgemeinschaft (SPP 1935) (to G.M.). Alexander Bio Phosphoimager (Dürr Medical). Data were quantified with ImageJ and Leitner thanks Ruedi Aebersold (Eidgenössiche Technische Hochschule Zürich) fitted with GraphPad Prism. for access to instrumentation and infrastructure. The Orbitrap Fusion Lumos mass spectrometer was purchased using funding from the Eidgenössiche Generation of a MED12 E33Q Knockin Mutant in HCT116 Cells. The HCT116 cell Technische Hochschule Scientific Equipment program and the European line carrying a MED12 E33Q knockin was generated using the CRISPR-Cas9 Union Grant ULTRA-DD FP7-JTI 115766 (to Ruedi Aebersold). 1. R. C. Conaway, S. Sato, C. Tomomori-Sato, T. Yao, J. W. Conaway, The mammalian 15. R. Firestein et al., CDK8 is a colorectal cancer oncogene that regulates β-catenin ac- Mediator complex and its role in transcriptional regulation. Trends Biochem. Sci. 30, tivity. Nature 455, 547–551 (2008). 250–255 (2005). 16. A. D. Clark, M. Oldenbroek, T. G. Boyer, Mediator kinase module and human tu- 2. S. Malik, R. G. Roeder, Dynamic regulation of pol II transcription by the mammalian morigenesis. Crit. Rev. Biochem. Mol. Biol. 50, 393–426 (2015). Mediator complex. Trends Biochem. Sci. 30, 256–263 (2005). 17. H. E. Pelish et al., Mediator kinase inhibition further activates super-enhancer- 3. J. Soutourina, Transcription regulation by the Mediator complex. Nat. Rev. Mol. Cell associated genes in AML. Nature 526, 273–276 (2015). Biol. 19, 262–274 (2018). 18. E. V. Schneider et al., The structure of CDK8/CycC implicates specificity in the CDK/ 4. B. L. Allen, D. J. Taatjes, The mediator complex: A central integrator of transcription. cyclin family and reveals interaction with a deep pocket binder. J. Mol. Biol. 412, 251– Nat. Rev. Mol. Cell Biol. 16, 155–166 (2015). 266 (2011). 5. M. T. Knuesel, K. D. Meyer, A. J. Donner, J. M. Espinosa, D. J. Taatjes, The human CDK8 19. M. Turunen et al., Uterine leiomyoma-linked MED12 mutations disrupt mediator- subcomplex is a histone kinase that requires Med12 for activity and can function associated CDK activity. Cell Rep. 7, 654–660 (2014). independently of mediator. Mol. Cell. Biol. 29, 650–661 (2009). 20. B. Nolen, S. Taylor, G. Ghosh, Regulation of protein kinases; controlling activity 6. Y. J. Kim, S. Björklund, Y. Li, M. H. Sayre, R. D. Kornberg, A multiprotein mediator of through activation segment conformation. Mol. Cell 15, 661–675 (2004). transcriptional activation and its interaction with the C-terminal repeat domain of 21. T. Dale et al., A selective chemical probe for exploring the role of CDK8 and CDK19 in RNA polymerase II. Cell 77, 599–608 (1994). human disease. Nat. Chem. Biol. 11, 973–980 (2015). 7. C. J. Hengartner et al., Association of an activator with an RNA polymerase II holo- 22. A. Leitner, M. Faini, F. Stengel, R. Aebersold, Crosslinking and mass spectrometry: An enzyme. Genes Dev. 9, 897–910 (1995). integrated technology to understand the structure and function of molecular ma- 8. C. Plaschka et al., Architecture of the RNA polymerase II-Mediator core initiation chines. Trends Biochem. Sci. 41,20–32 (2016). complex. Nature 518, 376–380 (2015). 23. A. Leitner et al., Chemical cross-linking/mass spectrometry targeting acidic residues in 9. K.-L. Tsai et al., Mediator structure and rearrangements required for holoenzyme proteins and protein complexes. Proc. Natl. Acad. Sci. U.S.A. 111, 9455–9460 (2014). formation. Nature 544, 196–201 (2017). 24. A. Leitner et al., Probing native protein structures by chemical cross-linking, mass 10. H. Elmlund et al., The cyclin-dependent kinase 8 module sterically blocks Mediator spectrometry, and bioinformatics. Mol. Cell. Proteomics 9, 1634–1649 (2010). interactions with RNA polymerase II. Proc. Natl. Acad. Sci. U.S.A. 103, 15788–15793 25. G. Lolli, Structural dissection of cyclin dependent kinases regulation and protein (2006). recognition properties. Cell Cycle 9, 1551–1561 (2010). 11. Z. C. Poss et al., Identification of Mediator kinase substrates in human cells using 26. S. Hoeppner, S. Baumli, P. Cramer, Structure of the mediator subunit cyclin C and its cortistatin A and quantitative phosphoproteomics. Cell Rep. 15, 436–450 (2016). implications for CDK8 function. J. Mol. Biol. 350, 833–842 (2005). 12. F. C. Holstege et al., Dissecting the regulatory circuitry of a eukaryotic genome. Cell 27. K. Huang et al., Structure of the Pho85-Pho80 CDK-cyclin complex of the phosphate- 95, 717–728 (1998). responsive signal transduction pathway. Mol. Cell 28, 614–623 (2007). 13. M. T. Knuesel, K. D. Meyer, C. Bernecky, D. J. Taatjes, The human CDK8 subcomplex is 28. C. Tarricone et al., Structure and regulation of the CDK5-p25(nck5a) complex. Mol. a molecular switch that controls Mediator coactivator function. Genes Dev. 23, 439– Cell 8, 657–669 (2001). 451 (2009). 29. P. D. Jeffrey et al., Mechanism of CDK activation revealed by the structure of a 14. L. D. Daniels et al., Mutual exclusivity of MED12/MED12L, MED13/13L, and CDK8/19 cyclinA-CDK2 complex. Nature 376, 313–320 (1995). paralogs revealed within the CDK-Mediator kinase module. J. Proteomics Bioinform. 30. N. P. Pavletich, Mechanisms of cyclin-dependent kinase regulation: Structures of Cdks, 1,1–7 (2013). their cyclin activators, and Cip and INK4 inhibitors. J. Mol. Biol. 287, 821–828 (1999). Klatt et al. PNAS Latest Articles | 11 of 12 Downloaded by guest on September 27, 2021 31. M. J. Park et al., Oncogenic exon 2 mutations in Mediator subunit MED12 disrupt 46. J. Bancerek et al., CDK8 kinase phosphorylates transcription factor STAT1 to selec- allosteric activation of cyclin C-CDK8/19. J. Biol. Chem. 293, 4870–4882 (2018). tively regulate the interferon response. Immunity 38, 250–262 (2013). 32. B. A. Schulman, D. L. Lindstrom, E. Harlow, Substrate recruitment to cyclin-dependent 47. B. Kroczynska, S. Mehrotra, A. D. Arslan, S. Kaur, L. C. Platanias, Regulation of kinase 2 by a multipurpose docking site on cyclin A. Proc. Natl. Acad. Sci. U.S.A. 95, interferon-dependent mRNA translation of target genes. J. Interferon Cytokine Res. – 10453 10458 (1998). 34, 289–296 (2014). 33. A. A. Russo, P. D. Jeffrey, A. K. Patten, J. Massagué, N. P. Pavletich, Crystal structure of 48. A. Subramanian et al., Gene set enrichment analysis: A knowledge-based approach the p27Kip1 cyclin-dependent-kinase inhibitor bound to the cyclin A-Cdk2 complex. for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U.S.A. 102, Nature 382, 325–331 (1996). 15545–15550 (2005). 34. K. Kämpjärvi et al., Somatic MED12 mutations are associated with poor prognosis 49. A. Liberzon et al., The Molecular Signatures Database (MSigDB) hallmark gene set markers in chronic lymphocytic leukemia. Oncotarget 6, 1884–1888 (2015). – 35. H.-R. Heinonen et al., Multiple clinical characteristics separate MED12-mutation- collection. Cell Syst. 1, 417 425 (2015). positive and -negative uterine leiomyomas. Sci. Rep. 7, 1015 (2017). 50. T. A. Wallace et al., Tumor immunobiological differences in prostate cancer between – 36. M. Yoshida et al., Frequent MED12 mutations in phyllodes tumours of the breast. African-American and European-American men. Cancer Res. 68, 927 936 (2008). Br. J. Cancer 112, 1703–1708 (2015). 51. M. Smid et al., Subtypes of breast cancer show preferential site of relapse. Cancer Res. 37. N. Mäkinen et al., MED12, the mediator complex subunit 12 gene, is mutated at high 68, 3108–3114 (2008). frequency in uterine leiomyomas. Science 334, 252–255 (2011). 52. K.-L. Tsai et al., A conserved Mediator-CDK8 kinase module association regulates 38. P. Bergeron et al., Design and development of a series of potent and selective type II Mediator-RNA polymerase II interaction. Nat. Struct. Mol. Biol. 20, 611–619 (2013). inhibitors of CDK8. ACS Med. Chem. Lett. 7, 595–600 (2016). 53. L. Jin, W. Wang, G. Fang, Targeting protein-protein interaction by small molecules. 39. C. Pargellis et al., Inhibition of p38 MAP kinase by utilizing a novel allosteric binding Annu. Rev. Pharmacol. Toxicol. 54, 435–456 (2014). site. Nat. Struct. Biol. 9, 268–272 (2002). 54. S. Huang et al., MED12 controls the response to multiple cancer drugs through reg- 40. E. V. Schneider, J. Böttcher, R. Huber, K. Maskos, L. Neumann, Structure-kinetic re- ulation of TGF-β receptor signaling. Cell 151, 937–950 (2012). lationship study of CDK8/CycC specific compounds. Proc. Natl. Acad. Sci. U.S.A. 110, 55. A. Leitner et al., Expanding the chemical cross-linking toolbox by the use of multiple – 8081 8086 (2013). proteases and enrichment by size exclusion chromatography. Mol. Cell. Proteomics 41. C. Alarcón et al., Nuclear CDKs drive Smad transcriptional activation and turnover in 11, M111.014126 (2012). – BMP and TGF-beta pathways. Cell 139, 757 769 (2009). 56. J.-B. Renaud et al., Improved genome editing efficiency and flexibility using modified 42. A. J. Donner, S. Szostek, J. M. Hoover, J. M. Espinosa, CDK8 is a stimulus-specific oligonucleotides with TALEN and CRISPR-Cas9 nucleases. Cell Rep. 14, 2263–2272 positive coregulator of p53 target genes. Mol. Cell 27, 121–133 (2007). (2016). 43. M. V. Dannappel, D. Sooraj, J. J. Loh, R. Firestein, Molecular and in vivo functions of 57. Y. Perez-Riverol et al., The PRIDE database and related tools and resources in 2019: the CDK8 and CDK19 kinase modules. Front. Cell Dev. Biol. 6, 171 (2019). – 44. F. Castro, A. P. Cardoso, R. M. Gonçalves, K. Serre, M. J. Oliveira, Interferon-gamma at Improving support for quantification data. Nucleic Acids Res. 47, D442 D450 (2019). the crossroads of tumor immune surveillance or evasion. Front. Immunol. 9, 847 58. M. J. Graham, C. Combe, L. Kolbowski, J. Rappsilber, xiView: A common platform for (2018). the downstream analysis of crosslinking mass spectrometry data. bioRxiv:10.1101/ 45. Z. Zhang, W. E. Theurkauf, Z. Weng, P. D. Zamore, Strand-specific libraries for high 561829 (26 February 2019). throughput RNA sequencing (RNA-Seq) prepared without poly(A) selection. Silence 3, 59. A. Lamiable et al., PEP-FOLD3: Faster de novo structure prediction for linear peptides 9 (2012). in solution and in complex. Nucleic Acids Res. 44, W449–W454 (2016). 12 of 12 | www.pnas.org/cgi/doi/10.1073/pnas.1917635117 Klatt et al. Downloaded by guest on September 27, 2021